(with David Fleet)
We propose a Bayesian framework for representing and recognizing local image motion in terms of two primitive models: translation and motion discontinuity. Motion discontinuities are represented using a non-linear generative model that explicitly encodes the orientation of the boundary, the velocities on either side, the motion of the occluding edge over time, and the appearance/disappearance of pixels at the boundary. We represent the posterior distribution over the model parameters given the image data using discrete samples. This distribution is propagated over time using the Condensation algorithm. To efficiently represent such a high-dimensional space we initialize samples using the responses of a low-level motion discontinuity detector.
M. J. Black and D. J. Fleet, Probabilistic Detection and Tracking of Motion Boundaries Int. J. of Computer Vision, 38(3):231-245, July 2000. (pdf).
Black, M. J. and Fleet, D. J., Probabilistic detection and tracking of motion discontinuities, Int. Conf. on Computer Vision, ICCV-99, Corfu, Greece, Sept. 1999, pp. 551-558. (postscript, 2.2MB) (Marr Prize honorable mention)